Principal Software Developer

PowerPlan • Atlanta, GA

2d • Hybrid

About The Position

As Principal Software Developer on PowerPlan's NXT platform, you will own delivery of scalable, maintainable, cloud-native services and help modernize a long-established platform used by regulated utilities. The platform is moving from legacy PowerBuilder systems to a modern C#/Angular architecture on Azure. You will contribute to architecture decisions and help shape engineering culture through standards for code, testing, and collaboration.

The role contributes to a significant platform transformation: reimagining proven functionality on a modern SaaS foundation while incorporating AI into product development and product capabilities. You will help deliver production-grade classical machine learning pipelines, RAG-powered search, LLM-driven assistants, prediction models, anomaly detection, and intelligent document processing.

Success involves becoming fluent in the NXT architecture, consistently shipping production-ready features with meaningful automated tests, advancing modernization and PowerBuilder deprecation, and making AI a dependable multiplier for engineering. You will also improve reliability, measurable productivity, and reusable tools and patterns for integrating ML models and LLM services. This is a hybrid role combining onsite work at the corporate office with work from home; onsite attendance may be required for scheduled office days, team or client meetings, and special events.

Requirements

• 7–10+ years of hands-on software development experience with enterprise-grade applications.
• Strong proficiency in C#/.NET, Angular or React, and cloud-native development on Azure.
• Experience modernizing legacy systems, including migrating from monoliths to modular services.
• Deep understanding of microservices, RESTful API design, and event-driven architectures.
• Working knowledge of CI/CD pipelines, automated testing, and DevOps practices.
• Working experience integrating classical machine learning models into application workflows, including prediction endpoints, feature pipelines, model serving (such as ML.NET, Azure ML, or ONNX), and feedback loops to improve model accuracy.
• Hands-on experience integrating Generative AI and LLMs, including RAG architectures using vector databases and embedding models, conversational features such as chatbots or copilot-style assistants, and prompt engineering with guardrails for enterprise and regulated environments.
• Familiarity with ML Ops practices, including model versioning, model A/B testing, monitoring model drift, and managing production inference pipelines.
• Experience using AI-assisted development tools such as GitHub Copilot, Cursor, or Windsurf, and familiarity with LLM-based tooling.
• Strong communication skills and ability to collaborate across architecture, product, QA, and CloudOps.
• An ownership mindset, including following through, escalating blockers early, and raising engineering standards.

Responsibilities

• Become fluent in the NXT architecture, assigned modules, legacy dependencies, cloud-native components, and current or planned ML, LLM, and data pipeline components within the first 4–6 weeks, then present findings to the Director of Software Development.
• Within the first six weeks, create a roadmap covering technical upskilling, backlog contribution targets, test coverage goals, at least three team tooling or workflow improvements, and a plan for deepening applied ML and GenAI skills; review it with the Software Development Manager.
• Ship production-ready features on an ongoing basis that meet NXT architecture standards, include meaningful automated tests, and require minimal rework.
• Deliver AI-powered capabilities, including prediction models, anomaly detection, RAG-based search, LLM-driven chatbots, and intelligent document processing.
• Modernize services, simplify legacy integrations, make measurable progress toward PowerBuilder deprecation, and identify opportunities for ML predictions or GenAI assistants to replace manual work or create product value.
• Within six months, demonstrate measurable productivity gains from AI-assisted development through improved cycle time, higher test coverage, or increased story throughput.
• Create reusable patterns, SDKs, or internal libraries that help teammates integrate ML models and LLM services into features.
• Maintain dependable ownership through strong unit, integration, and regression test coverage, meeting incident response SLAs, and reducing defects and interruptions over time.
• For AI features, monitor model performance, keep inference latency within SLA, and validate outputs against quality and safety thresholds, particularly in regulated financial and tax contexts.

What This Job Offers

• Job Type: Full-time
• Career Level: Principal
• Education Level: No Education Listed